We propose to study the impact of loss-of-function mutations of human DNA DSB (double-strand break) repair genes on rAAV (recombinant adeno-associated virus) vector integration. rAAV vectors have two powerful applications and they are in wide use in gene therapy studies. First, they can be used to deliver transgenes to cells. This aspect of rAAV, however, is plagued by the same insertional mutagenesis problems associated with other vectors. A second application for rAAV, however, is the recombination of the vector genome with its cognate homologous chromosomal sequences (aka, gene targeting) using HR (homologous recombination), which is a highly desirable outcome for basic researchers in need of tools to modify the genome and of great utility to gene therapists. However, because NHEJ (non-homologous end joining - the major pathway for DNA DSB repair - predominates in human cells over HR, random rAAV integration events generally occur much more frequently than correct gene targeting events. Therefore, therapeutic inactivation of the NHEJ pathway should augment rAAV-mediated gene targeting and - reciprocally - inactivation of the HR pathway should hinder rAAV gene targeting. In recently published work and in unpublished data we have demonstrated that reducing certain NHEJ factors does indeed increase the frequency of rAAV-mediated gene targeting. In this application we propose to extend this work to other NHEJ factors and to HR-regulated integrations. Moreover, we propose models - and experiments designed to test them - that should illuminate the mechanism that rAAV uses for gene targeting.